The 12-lead electrocardiogram as a biomarker of biological age

Abstract

Background We have demonstrated that a neural network is able to predict a person’s age from the electrocardiogram (ECG) [artificial intelligence (AI) ECG age]. However, some discrepancies were observed between ECG-derived and chronological ages. We assessed whether the difference between AI ECG and chronological age (Age-Gap) represents biological ageing and predicts long-term outcomes. Methods and results We previously developed a convolutional neural network to predict chronological age from ECGs. In this study, we used the network to analyse standard digital 12-lead ECGs in a cohort of 25 144 subjects ≥30 years who had primary care outpatient visits from 1997 to 2003. Subjects with coronary artery disease, stroke, and atrial fibrillation were excluded. We tested whether Age-Gap was correlated with total and cardiovascular mortality. Of 25 144 subjects tested (54% females, 95% Caucasian) followed for 12.4 ± 5.3 years, the mean chronological age was 53.7 ± 11.6 years and ECG-derived age was 54.6 ± 11 years (R2 = 0.79, P < 0.0001). The mean Age-Gap was small at 0.88 ± 7.4 years. Compared to those whose ECG-derived age was within 1 standard deviation (SD) of their chronological age, patients with Age-Gap ≥1 SD had higher all-cause and cardiovascular disease (CVD) mortality. Conversely, subjects whose Age-Gap was ≤1 SD had lower all-cause and CVD mortality. Results were unchanged after adjusting for CVD risk factors and other survival influencing factors. Conclusion The difference between AI ECG and chronological age is an independent predictor of all-cause and cardiovascular mortality. Discrepancies between these possibly reflect disease independent biological ageing.

Electrocardiogram-Based Heart Age Estimation by a Deep Learning Model Provides More Information on the Incidence of Cardiovascular Disorders

“…Furthermore, they investigated the residual between ECG-age and chronological age and found the residual is an independent predictor of all-cause Abbreviations: CVD, cardiovascular disease; ECG, Electrocardiography; DLM, deep learning model; AMI, acute myocardial infarction; CAD, coronary artery disease; CHF, congestive heart failure; AF, atrial fibrillation; DM, diabetes mellitus; HTN, hypertension; CKD, chronic kidney disease; AI, artificial intelligence; SCORE, Systematic COronary Risk Evaluation; ECG-age, ECG-based heart age; BMI, body mass index; HR, hazard ratio; CI, conference interval; ROC, receiver operating characteristic; AUC, area under curve; eGFR, estimated glomerular filtration rate; TG, triglycerides; COPD, chronic obstructive pulmonary disease; K, potassium; Na, sodium; Cl, chloride; Ca, calcium; Alb, albumin; GLU, glucose; HbA1c, glycated hemoglobin; BUN, blood urea nitrogen; Cr, creatinine; WBC, white blood cell count; PLT, platelet; Hb, hemoglobin; AST, aspartate aminotransferase; ALT, alanine aminotransferase; TG, triglyceride; TC, total cholesterol; LDL, low-density lipoprotein cholesterol; HDL, high-density lipoprotein cholesterol. mortality and cardiovascular mortality (23). Currently, many research teams had pointed out the strength of mortality risk stratification using ECG-age (24)(25)(26).…”

“…A previous study used 774,783 patients to train a DLM for predicting the age of the patient, which confirmed the feasibility of age extraction from ECG ( 22 ). Furthermore, they investigated the residual between ECG-age and chronological age and found the residual is an independent predictor of all-cause mortality and cardiovascular mortality ( 23 ). Currently, many research teams had pointed out the strength of mortality risk stratification using ECG-age ( 24 – 26 ).…”

Heart age estimated using explainable advanced electrocardiography

“…Attia, et al, showed that by using a deep neural network (DNN) artificial intelligence (AI) technique, a patient’s chronological age could be predicted, and that if the difference between the predicted and actual age was small, prognosis was good 8 . When Heart Age by Attia et al’s technique was older than the chronological age, the risk of future death was increased 7 . This corresponds well to the findings in our study that ECG Heart Age increased with increasing burden of cardiovascular risk.…”

“…If standard 10-second ECG recordings could be used instead, the clinical impact might be enhanced. Moreover, artificial intelligence has been used to estimate ECG Heart Age using the 10-second resting 12-lead ECG 7,8,10 . However, artificial intelligence techniques are limited by their “black box” approach, whereby the clinician does not have transparency as to the exact source(s) of the changes in the ECG that can affect an ECG Heart Age or other output 18,19 .Therefore, the aim of the study was to predict 5-minute ECG Heart Age from measures available by 10-second 12-lead ECG, and to compare the 10-second ECG Heart Age to chronological age in healthy subjects, subjects with cardiovascular risk factors, and patients with established cardiovascular disease.…”

Practical Lessons on 12-Lead ECG Classification: Meta-Analysis of Methods From PhysioNet/Computing in Cardiology Challenge 2020

“…For example, ventricular fibrillation is predominantly observed in the aged (Iwami et al, 2003). Recent studies have shown that the difference between the chronological age and DNN-estimated age can be used as a predictor of mortality (Ladejobi et al, 2021;Lima et al, 2021).…”